July 27, 1998

Brain Drain

With AT&T's vast assets, Michael Armstrong can easily acquire companies. Now comes the hard part: holding on to the talent.

By Scott Woolley

Cover of Forbes July 98 Issue

IN 1968 JOHN MALONE quit AT&T and soon joined a small company based on a fledgling technology: cable television. Thirty years later AT&T brought him back, buying his company for a staggering $48 billion.

AT&T is no longer the regulated monopoly that it was 30 years ago, but it has a similar dilemma: a massive pile of assets and an institutional reluctance to threaten those assets with new technologies.

The talent exodus exemplified by John Malone continues. Nine senior executives have recently jumped ship (see table). The talent is just as restless a few levels down: 15,300 of the company's 62,000 middle managers took early retirement last month, 50% more than the buyout plan was designed to attract. C. Michael Armstrong, AT&T's chief executive of eight months, has a big problem on his hands.

"I left because you needed 17 approvals for a decision," says Tom Evslin, who founded and ran WorldNet, AT&T's 1-million-subscriber Internet access division. Now he runs ITXC Corp., which sells newfangled Internet telephone calls to telecommunications companies at wholesale.

Techies have been fleeing AT&T because they don't want to be attached to obsolescent technologies. In a world moving to packet switching, AT&T is very dependent on old-style circuit-switched networks (see A switch in time).

One such AT&T engineer, David Isenberg, posted a devastating critique of his then-employer on the Internet a year ago. The pamphlet, entitled "The Rise of the Stupid Network," chastised AT&-and other established telcos-for not ditching their older technology. He compared them to 19th-century sailing merchants who "responded to the threat of steam by inventing faster sailing ships."

Fed up when the AT&T brass ignored him, the brainy Isenberg left AT&T Laboratories in January and now spends most of his time freelancing for startups besieging the old Ma Bell with newer technology.

As AT&T's new boss, Armstrong needs to change this sluggish culture. But he faces the same dilemma: How to embrace new technologies without obsoleting a huge fixed investment. "People stuck with an existing system always find it harder to change," says Alex Mandl, who abandoned his place as heir apparent to former AT&T chief executive Robert Allen in order to run startup Teligent.

Talk to other participants in the AT&T brain drain and you get a similar view. "They're captives of their cash flow and the old pricing umbrella," says Joseph Nacchio, who ran AT&T domestic long distance operations and who now runs Qwest, a long distance company undercutting AT&T's rates with a newer, higher-capacity fiber network. "That's part of the reason why so many of the entrepreneurial types left."

Armstrong acquired TCI and Teleport Inc. to bridge the so-called last mile connecting its global network to customers' homes and businesses- connections it lost when the old Ma Bell was broken up. AT&T points out that much of the talent that built those companies has stayed on. But neither acquisition addresses the fact that the long-haul network itself is decrepit. Much of AT&T's network has been around for close to a decade. That would be spanking new in a steel mill, but it's not in telecom: Qwest's fiber-optic cables are all less than three years old.

Nor do these big deals offset rapidly falling prices, as competitors less weighed down with fixed investment undercut AT&T. Ma Bell's customers made 8.7% more calls last year than in 1996, but revenues from those calls were flat at $46.2 billion. Earnings, more than 90% coming from long distance, dipped 20% to $4.6 billion.

"Long distance will be at 5 cents per minute by the end of the year, and that will still be artificially high compared to costs," says Mark Bruneau, head of Renaissance Worldwide's telecom consulting practice. Compare that with the prices AT&T now charges, in which typical consumer discount plans run 10 cents to 15 cents a minute.

If anyone can deal with this mess it is Armstrong. He gets plaudits, even from competitors like Mandl and Nacchio, for trying to change the old utility mind-set. By splitting AT&T stock into separate tracking stocks, he may free it from its slavery to earnings-per-share targets. The new capital structure isolates a steady earnings stream in one section of the business; the rest of the company is free to make bigger, riskier investments.

Armstrong has the means to buy innovative companies and he is doing so with great boldness. Now the question is: Does he have the means to keep the innovators?

See the table

Date last modified: 12 Oct 98